techUK论文-行动中的道德：从白皮书到工作场所

How Practical Implementationof the UK AI White Paper’sEthics Principles can beAchieved Through Assuranceand Standard November 2024 About techUK techUK is a membership organisation launched in 2013 to champion the technology sector andprepare and empower the UK for what comes next, delivering a better future for people, society, theeconomy and the planet. It is the UK’s leading technology membership organisation, with around1,000 members spread across the UK. We are a network that enables our members to learn from eachother and grow in a way which contributes to the country both socially and economically. By workingcollaboratively with government and others, we provide expert guidance and insight for our membersand stakeholders about how to prepare for the future, anticipate change and realise the positivepotential of technology in a fast-moving world. About techUK’s Digital EthicsProgramme In an increasingly digital world, it’s important that technology is used to improve and enhance thequality of people’s everyday lives. Embedding ethical principles, such as transparency, accountabilityand explainability, into the creation of products, tools and services is essential for building publictrust and confidence in technology. techUK focuses on resolving some of the most difficult ethicalchallenges, to ensure tech works for people and responsible innovation can flourish. Contents Foreword04 Introduction06 Principle 1: Safety, Security and Robustness10 Principle 2: Appropriate Transparency and Explainability14 Principle 3: Fairness18 Principle 4: Accountability and Governance22 Principle 5: Contestability and Redress27 Appendix35 References36 Foreword Since 2017, techUK has played a pivotal role in convening discussions on digital ethicsthrough our annual Digital Ethics Summit. Over the years, the landscape of digital ethicshas evolved significantly, with the development of new ethical frameworks, establishmentof institutions like The Ada Lovelace Institute, and DSIT’s Responsible Technology AdoptionUnit, and the growing community dedicated to addressing the complex challenges ofdigital ethics. Recent developments in Artificial Intelligence (AI), including breakthroughs in Generative AI (genAI),global summits on AI governance, and emerging regulatory proposals, have brought ethicalconsiderations at the forefront of technology policy. A key milestone in the UK was the government’s2024 AI White Paper1, which introduced a principles-based, context-specific regulatory framework forAI, empowering the existing regulators to address AI-related risks within their respective sectors. Thegovernment now looks to build on this framework. techUK has welcomed this approach, as it aligns with our calls for a pro-innovation, coordinated, andcontext-specific approach to AI regulation. In this context, closer regulatory cooperation, ensuring that regulators are equipped with the necessary capabilities, and making sure the UK’s approach isinteroperable with international standards will be vital for ensuring the UK remains competitive andaligned with global AI developments. We would be remiss not to acknowledge that the UK’s approach exists within a broader global context– the European Union’sAI Actis set to enter into force on 2 February 2025, while the United Stateshas introducedThe Executive Order for AIand the National Institute of Standards and Technology(NIST)AI Risk Management Framework, offering voluntary guidelines. China’s strategy focuseson national security and technological independence, while the APAC region is developing diverseregulatory approaches. However, due to scope constraints, this paper will focus specifically on theUK’s AI White Paper approach. Our aim is to provide in-depth insights into how the UK’s principles-based framework can be effectively implemented by organisations operating within the UK regulatorylandscape. As the AI landscape continues to transform, organisations face a number of practical challengesin implementing AI ethical principles in real-world contexts. Balancing innovation with ethicalconsiderations, interpreting broad principles in specific contexts, addressing the shortage of qualifiedpersonnel to implement and interpret tools, ensuring consistent application across diverse projects,and keeping pace with rapidly evolving AI technologies are just a few difficulties they must navigate.These challenges underscore the need for a practical, actionable framework to guide businessesthrough ethical AI implementation. This paper recognises that regulatory guidelines are at times perceived as abstract or theoretical,while industry practices are seen as more concrete and applied. The aim of this paper is to bridgethis gap, demonstrating how the regulatory principles, set out in the AI White Paper, are not merelytheoretical constructs, but are already being operationalised within the industry. Through insights,illustrative tools, and real-world examples, the paper aims to of